A modestly hypofractionated course of chemoradiation (CRT) consisting of 36 Gy/15 fractions (F) concurrent with gemcitabine used in PREOPANC and phase II trials has become increasingly common for the treatment of borderline resectable (BR) and locally advanced (LA) pancreatic cancer (PC). Achieving an R0 resection remains a key prognostic factor in PC. We tested whether escalating dose beyond standard dosing (SD) of 36-39 Gy/15 F (or 50-54 Gy/25-30 F) would improve R0 resection rates and outcomes while respecting nearby organs at risk. This was a retrospective analysis of consecutive patients at our institution from 2012-2022 with BR/LA PC treated with moderate dose escalated (MDE) (45 Gy/15 F, N = 45) or SD (36-39 Gy/15 F, N = 68 or 50-54 Gy/25-30 F, N = 25) CRT. For MDE, a 5 mm expansion from the duodenum, small bowel, and stomach was created (GI_PRV); PTV was cropped from this structure and prescribed 45 Gy/15 F. The primary endpoint was R0 resection rate with secondary endpoints of cumulative incidence of local progression (LP, recurrence after surgery/imaging progression if no surgery) with death as a competing risk (LP after occurrence of distant metastasis [DM] were still captured), cumulative incidence of DM, and overall survival (OS). Univariable and multivariable competing risks regression analyses were performed to determine the association between baseline covariates and LP. We identified 45 patients treated with MDE and 93 treated with SD. Most patients presented with BR disease (55.6% MDE; 54.8% SD) and received neoadjuvant chemotherapy with FOLFIRINOX (98% MDE; 99% SD). All patients in the MDE group and 99% in the SD group received concurrent chemotherapy with gemcitabine used most often (96% MDE; 77% SD). Median follow-up was 17 m (IQR 13-27 m). Surgical resection rates were similar between groups (33.3% MDE vs. 39.8% SD, p = 0.46). Amongst patients that had surgery, R0 resection rates were non-significantly higher in the MDE group (73.3% vs. 47.4%, p = 0.09). Cumulative incidence of LP at 18 m was significantly lower in the MDE group (9.0% vs. 24.8%, p = 0.04). No difference in rates of DM (51.2% MDE vs. 59.6% SD, p = 0.92) or OS at 18 m (53.9% vs. 53.6%, p = 0.89) were observed. On multivariable analysis, MDE (HR = 0.39, p = 0.03) and pancreatic head location (HR = 0.51, p = 0.04) were the only factors independently associated with LP. Rates of grade 2+ gastrointestinal toxicity during CRT (20% MDE vs. 20.9% SD, p = 0.91) and ≤90 days of completing CRT (11.6% MDE vs. 14.8%, p = 0.62) were similar between groups, as were rates of grade 3+ hematologic toxicity (52.3% MDE vs. 41.3% SD, p = 0.23). In this single institutional study, we found MDE is a simple, safe, and effective strategy associated with improved local control, higher R0 resection rates, and similar toxicity to SD CRT for patients with BR/LA PC. Further prospective data is needed to clarify the role of dose-escalated RT in the management of this lethal malignancy.
Read full abstract